David adjusted his tie for the tenth time. The open house was about to start, and he felt a knot of anxiety tighten in his chest. He hadn't been able to sleep well since the investigator had informed him about the history of the house. The previous owner had been found dead in the master bedroom. Now, every creak of the floorboards, every shadow cast by the setting sun, amplified his unease.

The first couple through the door seemed pleasant enough, but David found himself fidgeting, pointing out trivial features as if desperately trying to distract himself. He kept glancing at the master bedroom door, willing it not to draw their attention. He couldn't help but imagine the faces of the people who knew the victim.

He tried to keep his voice steady as he described the newly renovated kitchen, but his heart was hammering against his ribs. He felt the blood drain from his face as a woman pointed to a stain on the wall and asked what it was. "Water damage," he blurted out, before quickly redirecting the conversation.

Emotion: self-conscious

Cluster: Embarrassment
PC1 (Valence): -1.03 Negative
PC2 (Disposition): 0.52

Role in Research

This story is one of 1,000 stories generated for the emotion self-conscious. During extraction, it was fed through Gemma4-31B and its hidden state activations were captured at 11 layers.

The mean activation across all 1,000 self-conscious stories, after denoising with neutral dialogue baselines, produces the self-conscious emotion vector -- a direction in the model's 5,376-dimensional representation space.

Logit Lens (Layer 40)

Tokens promoted/suppressed when the self-conscious vector is projected through the unembedding matrix.

Promoted:
S0.307
C0.298
ness0.273
尴尬0.263
😶0.242
Suppressed:
que-0.399
de-0.266
triumphant-0.266
আনন্দের-0.247
ايا-0.238